A kinetic model for predicting the oxidative degradation of additive free polyethylene in bleach desinfected water

https://doi.org/10.1016/j.polymdegradstab.2017.09.020Get rights and content

Highlights

  • Formation of Cl2, ClOH and ClO- in the water phase.

  • Migration of Cl2 and ClOH into the PE matrix.

  • Dissociation of Cl2 and ClOH into Cl and HO radicals.

  • Radical attack of the PE matrix initiating its oxidation.

Abstract

The chemical interactions between additive free PE and bleach were investigated by FTIR spectrophotometry and viscosimetry in molten state after immersion (for a maximum duration of one hundred days) in bleach solutions maintained at a temperature of 60 °C, a free chlorine concentration of 100 ppm, and a pH = 4, 5 or 7. It was found that the polymer undergoes a severe oxidation from the earliest days of exposure in a superficial layer of about 50–100 μm thick, almost independent of the pH value. In this layer, oxidation leads to the formation and accumulation of various carbonyl products (mostly ketones and carboxylic acids) but also, after about 2–3 weeks of exposure, to a dramatic decrease in the average molar mass due to the large predominance of chain scissions over crosslinking. It was also found that the oxidation rate is maximum at pH = 5, and of the same order of magnitude at pH = 4 and 7. Based on the equilibrium diagram giving access to the relative predominance of the three main chemical species as a function of the pH value of the bleach solution, it was assumed that oxidation is initiated by radical species coming firstly from hypochlorous acid (ClOH) and secondarily from chlorine (Cl2), given that hypochlorite ions (ClO) are totally insoluble into the PE matrix. In addition, for explaining the surprisingly large value of the oxidized layer thickness despite the high reactivity of the involved radicals, it was assumed that ClOH and Cl2 do not decompose into radicals in the water phase, but migrate deeply into the PE matrix prior to dissociating into Clradical dot and HOradical dot radicals and then, initiating a radical chain oxidation. The validity of the kinetic model derived from this scenario was successfully checked by comparing the numerical simulations with all the experimental data collected in this study. This model predicts the general trends of the oxidation kinetics and its dependence on the pH value, but also gives access to the transport properties of the chlorinated disinfectants and their radical species, and the rate constants of the radical attack.

Introduction

Steel pipes (with or without polymeric coating) are traditionally used for tertiary cooling circuits in nuclear power plants. Water is commonly disinfected by chlorinated reagents in order to control the microbiological growth and limit the deposit and development of shellfish and seaweed that might clog the circuit input. However, steel is sensitive to corrosion and water disinfectants are relatively strong oxidizers susceptible to accelerate corrosion. Therefore, in recent years, the nuclear industry has launched heavy testing campaigns for qualifying polymeric materials for this type of application, in particular polyethylene (PE), in view of using them in replacement of steel.

PE pipes are currently used for conveying drinking water under a pressure of a few bars since the early 1970s. There is an abundant literature on results of isobaric and isothermal ageing tests in pure distilled water at temperatures higher than typically 40 °C (see for instance references [1], [2], [3]). The construction of a single master curve by using adequate Arrhenius shift factors [2] allows extrapolating these results up to ambient temperature and thus, demonstrating that pipes perish always by brittle fracture with lifetimes exceeding typically 50 years in the domain of water pressures of practical interest.

In the past half century, considerable research efforts were accomplished for optimizing the polymer structure, at both macromolecular and morphological scales [4], [5], [6], [7], [8], [9], [10], but also the processing conditions [11] in view of improving the pipe durability. In the last two decades however, it was discovered that the water disinfectants do not only destroy the organic substances in water by oxidizing processes (among which radical processes play a key role), but also consume the stabilizing function of phenolic antioxidants and initiate a radical chain oxidation of the polymer matrix in the inner wall of pipes, thus leading to a significant reduction the pipe lifetime. These deleterious effects are especially pronounced when chlorine dioxide is used as a disinfectant [12], [13], [14], but measurable effects were also evidenced in the case of chlorine and bleach [15], [16], [17], [18], [19], [20], [21].

The chemistry of water disinfection is relatively well known. Both radical and ionic species (for instance, ClO in the case of chlorine and bleach) have a strong oxidizing power. However, PE acts as a selective absorber because highly polar species such as ions are totally insoluble into this non-polar matrix. This characteristic was first discovered by Ravens [22] in the case of the poly(ethylene terephthalate) (PET) hydrolysis. PE is part of the less polar polymers: its dipolar moment is zero and its dielectric permittivity is 2.3 [23]. In contrast, radicals are more or less soluble into PE depending on their solubility parameter. As a result, three distinct scenarios can be considered for explaining the chemical attack of both phenolic antioxidants and PE matrix:

  • -

    S1) The disinfectant is itself a free radical in ground state which can migrate into PE. This is the case of chlorine dioxide (ClO2) [12], [13];

  • -

    S2) The disinfectant generates radicals in water, then these latter migrate into PE;

  • -

    S3) The disinfectant itself, or a non-dissociated molecule formed from this disinfectant in water, migrates deeply into PE where it dissociates into radicals.

The identification of the reactive species and the determination of their respective concentration in the PE matrix remain an open issue, especially in the case of chlorine and bleach for which the two last scenarios S2 and S3 are conceivable. It is now well known that bleach solutions contain three main chemical species whose the relative proportions depend essentially on the pH value [24], as shown for instance at 25 °C in Fig. 1. These latter can be determined from the kinetic analysis of the two following chemical equilibria:Cl2 (aq) + 2H2O ⇆ ClOH + Cl + H3O+ClOH + 2H2O ⇆ ClO + H3O+

It comes:%Cl2=100×10pH+pKd1+10pH+pKd+10pHpKa%ClOH=1001+10pH+pKd+10pHpKa%ClO=100×10pHpKa1+10pH+pKd+10pHpKa

The temperature dependence of the equilibrium constants Kd and Ka was reported in the literature [25], [26]. Typically between 0 and 45 °C, it can be written:pKd=982798T25485.7T+10.7484pKa=3000T10.0686+0.0253 T

Since Kd and Ka are very slowly decreasing functions of temperature, the chemical composition of bleach is almost insensitive to temperature. In fact, the chemical equilibria are very slightly shifted towards lower pH values when increasing the temperature. It can be thus concluded that chlorine (Cl2), hypochlorous acid (ClOH) and hypochlorite ion (ClO) predominate respectively in highly acidic (pH < 3), weakly acidic (3 < pH < 7.5) and basic media (pH > 7.5), whatever the temperature. Moreover, the maximum yield of ClOH is reached at pH = 5.

Based on these observations, it is thus possible to select carefully pH conditions for identifying what chemical species are responsible for the polymer degradation, but also for deciding between the two possible scenarii S2 and S3.

According to Holst [27], when they coexist (at pH = 7.5 ± 1.5), ClOH and ClO would generate ClOradical dot and HOradical dot radicals in water:ClOH + ClO → ClO + Cl + HOHOradical dot + ClO → HO + ClOradical dotClOradical dot + ClO + HO → 2Cl + O2 + HOradical dot

Thus, scenario S2 could be considered. This scenario was supported by some authors who effectively found a maximum degradation rate at pH = 7 for polyether-based polyurethane fibers [28], or at pH = 8 for polysulfone membranes [29], [30], [31].

On the contrary, for other authors [18], [20], [32], [33], Cl2 and ClOH could directly dissociate into Clradical dot and HOradical dot radicals (presumably within the PE matrix) because Cl−Cl and Cl−O bonds are characterized by a very low dissociation energy (of respectively 242 and 247 kJ mol−1 [34]):Cl2 → 2 Clradical dotClOH → Clradical dot + HOradical dot

In this case, the degradation rate would be maximum at pH < 7 and scenario S3 could be envisaged.

Regardless the chosen scenario, it will remain to determine the relative proportion of each radical involved in the polymer degradation events. Since HOradical dot radicals are of the smallest molecular size, they have a high diffusivity into the PE matrix, presumably very close to that of water [35]. However, as they are also extremely reactive with respect to hydrocarbon substrates (typically more than 1010 times more reactive than PO2radical dot radicals at 25 °C [36], [37], [38]), they should be scavenged into a sub-micrometric superficial layer of PE pipes. In contrast, chlorinated radicals (ClOradical dot and Clradical dot) should diffuse more slowly into the inner pipe wall because of their larger molecular size, but they should be also much less reactive than HOradical dot radicals. In fact, it is expected that their reactivity ranks in the following order:HOClOCl

In a first approach, the depth L of penetration of a radical species can be estimated from a simple scaling law [39]:L(DK)1/2where D is the diffusion coefficient and K the first-order constant for the consumption of the diffusing species by the chemical reaction with the polymer.

As an example, this equation can be used to compare HOradical dot and ClOradical dot radicals:LHOLClO=(DHODClOKClOKHO)1/2

The diffusion coefficients of both radicals into the PE matrix are not known, but by analogy with more common vapors and gases of very close molecular size, for instance H2O [35] and CO2 [23], it can be written:1<DHODClO<10

So that, presumably:LHOLClO103102

Chlorinated radicals are thus expected to penetrate more deeply into the inner wall of pipes. The order of magnitude of the depth of this radical attack should also allow us to decide between scenarii S2 and S3.

The objectives of the present article are twofold. The first objective is to evidence and analyze by conventional laboratory techniques the chemical interactions between additive free PE and bleach in view of proposing a general degradation mechanism. Since ClOH is often considered as the main source of radicals in the literature [27], [28], [29], [30], [31], [32], [33], three distinct pH values, corresponding to the maximum yield (pH = 5) and lower but equivalent yields of ClOH (pH = 4 and 7), will be investigated. The two last values are necessary to see if Cl2 may be a secondary source of radicals. If so, the degradation rate will be less reduced in acidic (pH = 4) than in neutral medium (pH = 7). In these exposure conditions, two aging indicators will be used for deciding between scenarii S2 and S3: the oxidation rate and the thickness of the oxidation layer. The second objective is to derive a kinetic model from the degradation mechanism and to check its validity by comparing the numerical simulations with all the experimental data collected in this study.

Section snippets

Materials

An unstabilized and unfilled PE powder was supplied by Borealis Company for this study. PE films of thicknesses ranging between 150 and 350 μm were formed by compression molding with a Gibrite laboratory press under a pressure of 3 MPa for 2 min at 180 °C. After demolding, the films were characterized by conventional laboratory techniques.

The FTIR spectrophotometry was used in a transmission mode to check that the molding conditions had avoided a premature oxidation of the films. No trace of

Experimental results

The changes in the average carbonyl concentration [CO]global and weight average molar mass MW global of PE films are reported in Fig. 6 for all the exposure conditions under study. It appears clearly that the polymer undergoes a severe oxidation from the earliest days of exposure, but the average oxidation rate is faster at pH = 5 than at pH = 4 and 7. In addition, the kinetics curves reported at pH = 4 and 7 superimpose perfectly (if taking into account the measurement uncertainties). Based on

Conclusion

The chemical interactions between additive free PE and bleach have been investigated in solutions maintained at a temperature of 60 °C, a free chlorine concentration of 100 ppm, and a pH = 4, 5 or 7. A scenario was proposed for tentatively explaining why oxidation reaches its maximum rate at pH = 5 and occurs in a relatively large superficial thickness (about 50–100 μm thick), almost independent of the pH value, despite the high reactivity of the involved radicals (in particular HO). According

References (63)

  • X. Colin et al.

    Lifetime prediction of polyethylene in nuclear plants

    Nucl. Instrum. Methods Phys. Res. Sec B

    (2007)
  • X. Colin et al.

    Determination of thermal oxidation rate constants by an inverse method. Application to polyethylene

    Polym. Degrad. Stab.

    (2004)
  • G. Moad et al.

    Aqueous hydrogen peroxide-induced degradation of polyolefins: a greener process for controlled-rheology polypropylene

    Polym. Degrad. Stab.

    (2015)
  • M. Ifwarson

    Life-time of polyethylene pipes under pressure and exposure to high temperatures

    Kunstst. Ger. Plast.

    (1989)
  • Y. Lu et al.

    The ductile-brittle transition in a polyethylene copolymer

    J. Mater Sci.

    (1990)
  • H.-H. Kausch

    Polymer Fracture

    (1987)
  • P.G. De Gennes

    Reptation of a polymer chain in the presence of fixed obstacles

    J Chem Phys Am. Inst. Phys.

    (1971)
  • Y.L. Huang et al.

    Dependence of slow crack growth in polyethylene on butyl branch density: morphology and theory

    J. Polym. Sci. B Polym. Phys.

    (1991)
  • N. Brown et al.

    Slow crack growth in polyethylene - a review

    Makromol. Chem. Macromol. Symp.

    (1991)
  • T. Trankner et al.

    Molecular and lamellar structure of an extrusion-grade medium-density polyethylene for gas distribution

    Polym. Eng. Sci.

    (1994)
  • M.S. Hedenqvist

    Morphology and Morphology Sensitive Properties of Polyethylene: Fracture Behavior and Diffusivity

    (1995)
  • U.W. Gedde et al.

    Molecular structure, crystallization behavior, and morphology of fractions obtained from an extrusion grade high-density polyethylene

    Polym. Eng. Sci.

    (1988)
  • A.J. Peacock

    Handbook of Polyethylene. Structures, Properties and Applications

    (2000)
  • X. Colin et al.

    Aging of polyethylene pipes transporting drinking water disinfected by chlorine dioxide. I- Chemical aspects

    Polym. Eng. Sci.

    (2009)
  • X. Colin et al.

    Aging of polyethylene pipes transporting drinking water disinfected by chlorine dioxide. I- Lifetime prediction

    Polym. Eng. Sci.

    (2009)
  • W. Yu et al.

    Deterioration of polyethylene pipes exposed to water containing chlorine dioxide

    Polym. Degrad. Stab.

    (2001)
  • S.W. Bradley et al.

    The effect of chlorine on the long-term durability of crosslinked polyethylene pipe

  • M. Ifwarson et al.

    Results and experiences from tests on polyolefin pipes exposed to chlorinated water

  • T.S. Gill et al.

    Long-term durability of crosslinked polyethylene tubing used in chlorinated hot water systems

    Plast. Rubb Compos

    (1999)
  • J.P. Dear et al.

    The effects of chlorine depletion of antioxidants in polyethylene

    Polym. Polym. Compos

    (2001)
  • J.P. Dear et al.

    Effect of chlorine on polyethylene pipes in water distribution networks

    J. Mater Des. App

    (2006)
  • Cited by (35)

    • Gravimetric analysis of stability of polymeric materials during exposure to chemical disinfectants at different temperatures

      2022, Chemosphere
      Citation Excerpt :

      Chlorine dioxide (ClO2) and hypochlorite (HOCl) are the most commonly used chemical disinfectants in water distribution systems for municipal water supply to deactivate microorganisms and other contaminants (Connick and Chia, 1959; Morris, 1966; Mikdam et al., 2017; Khan et al., 2019, 2020b).

    • Modeling of multiple crack initiation in polymer pipes under oxidative environment

      2022, International Journal of Engineering Science
      Citation Excerpt :

      In particular, this failure mode should be carefully considered for potable water transportation systems, where water is commonly disinfected to remove bacteria and harmful germs using chlorine-based disinfectants. Because chlorine-based disinfectants are strong oxidizers, they can quickly deplete antioxidants (AO) and deteriorate the polymer substance, which can eventually lead to unexpected cracking (Mikdam et al., 2017; Yu et al., 2011, 2015). In addition, recent attempts have been made to apply high-density polyethylene pipes as power plant piping because they have superior structural integrity under displacement-controlled loads, such as outstanding resistance to seismic loads and soil movement, compared with carbon steel (Kalyanam et al., 2011; Luo et al., 2015, 2014; Sheng et al., 2017; Shi et al., 2021; Zheng et al., 2018).

    • Degradation analysis of polymeric pipe materials used for water supply systems under various disinfectant conditions

      2022, Chemosphere
      Citation Excerpt :

      The chain secession was prominent in the 10-mg/L doses of the ClO2 and NaOCl as the C–C/C–H chain loss resulted in increased R-CO-OR groups, confirming the formation of esters, and carboxylic acid groups were confirmed through the ART-FTIR analysis, as explained earlier. Furthermore, the decrease in peak area of ketones and aldehydes showed that, in the initial stages, the pipe stabilizers/antioxidants were removed to form the quinoid structure, which allowed the disinfectant to penetrate into the main polymer (Dear and Mason, 2001; Mikdam et al., 2017) and allow deterioration to form the R-CO-OR groups. As a result, rapid degradation was observed to form the carboxylate groups, more prominently observed in the LDPE and Hi-PVC cases.

    • Chemo-mechanical modeling of static fatigue of high density polyethylene in bleach solution

      2021, International Journal of Solids and Structures
      Citation Excerpt :

      Mikdam et al. (2017) developed a corrosion kinetics model for polyethylene in bleach disinfected water. This model predicts the time evolution and space distribution of the concentration of reactants, the number of chain scission and cross-linking events, as well as the concentrations of bi-products (Mikdam et al., 2017). Hereafter we refer to Mikdam’s model as the full model.

    • A new analytical model for predicting the radio-thermal oxidation kinetics and the lifetime of electric cable insulation in nuclear power plants. application to silane cross-linked polyethylene

      2021, Polymer Degradation and Stability
      Citation Excerpt :

      At this stage of investigations, it can be seen that the three analytical equations allow accounting for the general trends of the oxidation kinetics under these critical exposure conditions (i.e. under low dose rate at moderate temperature) where, obviously, a stronger coupling between radiochemical and thermal oxidations operates. The orders of magnitude previously determined on linear PE (i.e. LDPE, MDPE and HDPE) for almost all these kinetic parameters were kept for these simulations [1-3,20,36,39,43,44]. Only k5 was slightly reduced at low temperature close to ambient (i.e. at 47 and 21°C) because it had already been proposed to allocate a small activation energy to this termination rate constant in order to better simulate the oxidation kinetics initiated by chlorine disinfectants in PE pipes carrying drinking water at ambient temperature [20,43,44].

    View all citing articles on Scopus
    View full text